The Average Discord Server Has 883 Members. The Typical One Has 30.

If you run a Discord server with 80 members, you have wondered whether that is any good.

We studied 211,039 real communities to find out. Here is the answer: you are bigger than roughly 70% of them.

The number everyone quotes is the average. The average is 883 members. That number is real and it is also garbage, because Discord server sizes do not cluster around a middle - a handful of servers have millions of members and most have a couple of dozen. The mean Discord server is larger than roughly 90% of Discord servers.

The actual middle is 30.

Which means almost everything you believe about your own server is calibrated against outliers. Below is where you really rank - and once the numbers are out of the way, the five changes that move you up the fastest.

Half of Discord is smaller than a school class#

Half of all Discord communities have fewer than 100 members, and three quarters have fewer than 1,000. Only about one in a thousand has more than 100,000. That is the whole shape of the platform in three numbers, and it looks nothing like the impression you get from the servers people talk about. Large communities are visible precisely because they are large: they get recommended, screenshotted and linked, while the tens of thousands of servers with forty regulars are invisible by nature. So the reference point almost every owner carries around is built entirely from outliers, which is why so many perfectly healthy servers feel like failures to the people running them. Measured against the real distribution, a server with a few dozen active people is not a failed community. It is the most common kind there is, and usually the kind worth being in.

One quick win while you are here: if your server has more than four channels and fewer than 200 members, you have already found your biggest problem. More on why in a moment.

Where you rank, exactly#

Read that table again, because it is not what anyone expects. One hundred members puts you in the top 25% of every Discord community there is. Three hundred and sixty-nine puts you in the top 10%. The jump from there to 8,793 covers the last nine percent, which is where all the servers you have heard of live.

So forget ten thousand. The target that actually means something is closer than you think.

The mistake that makes 40 members look like 4#

Here is what nobody tells you when you start a server: your channel list is doing more damage than your member count.

A new owner builds the server they hope to have. General, off-topic, memes, introductions, announcements, voice, art, music, bot-commands. Eight rooms. Then forty people join, spread across all eight, and every single one looks empty - because forty people divided by eight channels is five people per room, and five people across a week is silence.

The same forty in two channels look like a community.

So delete channels. You can add them back the day a conversation genuinely splits, which is a good problem and will not happen for months. It is the highest-leverage change available to a server under 200 members and it takes about ninety seconds.

That is one of five. The other four are below - and one of them matters more than the rest combined, though it is not the one people expect.

The four that actually move the number#

Invite ten people you already know, before any strangers. An empty server converts nobody. Ten friends who will genuinely post give the eleventh person a reason to stay, and the eleventh is the first one you did not have to recruit.

Post first, every day, for a month. Early on you are not the moderator, you are the first guest. A server where the owner posts daily reads as alive. The identical server where they post weekly reads as abandoned. Same members, completely different impression.

Give people one reason to return on a specific day. Anything repeating beats anything one-off - a Thursday game night, a weekly thread, a standing prompt. The mechanism is simple: it turns a server I joined once into a thing I do.

Spend the effort on ten people, not a thousand. When the count feels slow the tempting move is bought members or mass invites, and it backfires in a specific way: silent accounts make your twenty real ones look like nobody is home, so the next genuine visitor leaves. The inverse is the actual cheat code. Ten people who feel needed - one runs the game night, one posts the prompt, one greets new joins - will build a server that runs without you. That is worth more than a thousand names, and it is the thing every large server on this list started with.

The number worth remembering#

30 members is the middle of Discord. Not the floor, not a disappointment. The middle.

More than that and you are in the top half of every community on the platform. Fewer and you are in the phase every large server passed through - including all 218 of the ones with six-figure member counts, every one of which was once four people and an empty general channel.

The genuinely useful reframe hiding in this data: you are not competing with the servers you have heard of. You are competing with the median, and the median is thirty people. A server with thirty regulars who actually talk is already better than most things on this platform - it just does not look like it from the inside, because the sidebar counts the wrong thing.

The other half of the picture is time, and we measured that separately: how long it actually takes to grow a Discord server. Short version - a server started this year typically has 13 members, and the curve does something unexpected after year one.

Want to see where you sit against real communities? Browse active gaming servers, anime communities and social servers on Rally. Member counts are on every listing.